( - )-7-HYDROXY-P-ISOSPARTEIKE
OCTOBER1967
3045
( - )-7-Hydroxy-p-isosparteine, an Alkaloid Accompanying p-Isosparteine in Lupinus sericeus Purshl' J f A R V I N CARM.4CK,lb S T A N L E Y
I. GOLD BERG,'^
ASD
ERICw.> 1 A R T I N 2
Gdpartment of Chemistry, Indiana University, Bloomington,Indiana 47401, and Department of Chemistry, Cniversify of Pennsylvania, Philadelphia, Pennsylvania Received September 30, 1966
A crystalline alkaloid (Cl5H&2O) which accompanies ( - )-p-isosparteine3-5in specimens of the plant, Lupinus sericeus Pursh, grown in southern Utah, has been shown by chemical evidence to be either 'i-hydroxy-p-isosparteine or the 8-hydroxy isomer. The 7-hydroxy structure is favored by the chemical evidence and independently confirmed by X-ray crystallographic investigation.6
Specimens of dried, above-ground parts of the plant, Lupinus sericeus Pursh, collected near Salina, Utahj5v7 yielded ( -)-p-isosparteine as the principal alkaloid. The chemical proof of its structure and stereochemical relationship to (+) -sparteine have been previously described. 3-5 The absolute configuration of (- )-p-isosparteine (I) follows from our earlier work4 showing the identity of the configurations of C-7 and C-9 in (-)-p-isosparteine and in (+)-sparteine, and from the interconversions of Okuda, Tsuda. and Iiataoka.*
I
11
(1) (a) Contribution No. 1482 from the Department of Chemistry, Indiana University. (b) T o whom inquiries should be address: Department of Chemistry, Indiana University, Bloomington, Ind. 47401. (e) h portion of the experimental v o r k is from t h e Ph.D. Dissertation of S. I. Goldberg, Indiana University Graduate School, 1958; Chem. Ahstr., 63, 3257~' (19.39). (2) T h e original isolation of t h e two principal alkaloids of L. sericeus Pursh was described in t h e 1'h.D. Dissertation of E. \V. Martin, submitted t o the Graduate Srhool, University of Pennsylvania, 1949. Regarding names. cf. footnote 3. (3) \Ve have adopted the name, B-isosparteine, t o describe the principal alkaloid of L. serzceus Pursh a n d discontinue t h e use of t h e name, spartalupine, used in previous papers,Z'h a s recommended by Leonard5 There no longer appears t o be likelihood of confusion of this alkaloid with t h e substance described a s B-isosparteine by K. Kinterfeld and C . Rauch, 4 r c h . P h a r m . , 272, 273 (1934). (4) (a) &I. Carmack and. E. \I-, Martin. Abstracts of Papers, 124th National Meeting of the American Chemical Society, Chicago, Ill.. Sept 1953, 32-0. (b) 21. Carmack, B. Douglas, E. K. Martin, and Hanna Suss, J . Am. C h e m . SOC.,77, 4136 (1955). (5) N. J. Leonard, "The Alkaloids. Chemistry and Physiology," Vol. VII, R. H. F. J f a n s k e , Ed., Academic Press Inc., N e w York, N. T.,1960, pp 274, 284, 285, 28!3. (6) J. M. H. Pinkcarton and L. K. Steinrauf, J . 070. Chem., 32, 1828 (1967). (7) Collections of plant material were made during t h e summers of 1913, 1947, and 1954 under the direction of Dr. W. T. Huffman. Veterinarian in Charge, Stock Poisoning by Planta, Salt Lake City Office, Bureau of Animal S. Department of Agriculture Research Administration, U. 6. Industry, Department of .igricuiture. T h e lupine plants were collected during tlie flowering seaqon in early eummer a t t h e Salina Experiment Station of t h e C. S. Department of Agriculture, in the Fishlake National Forest, 16 miles from Salina, C t a h . Botanical specimens from these collections were examined b y several botanists and were found to have characteristics of t h e species Lupinus spathulalus, L . ornntus, and L. sericeus Pursh. T h e latter classification, which Ire have adopted, was made by Charles Piper Smith, Cupertino, Calif., on the basis of specimens from tlie collection made in 1947 by the Utah group. Smitli stated, "The specimen is clearly L. sericeus Pursh a s represented in t h e moiintains of Utah, a t least about Coulter, Soldier Summit, and Park City plus t h e upper Provo River valley-except t h a t the specimen here concerned has the upper surfaces of the leaflets subsericeiis t o glabrous for the cldest leaves. This one variation points to relationship with L. marianus Rydh., which has t h e leaves scattered on scattered stems, leaflets green and subsericeus below, glabrate or glahrous above." Although the yields varied slightly, t h e same alkaloids were found in all three collections made in different years. T h e study of the alkaloids ma)- prove helpful in the solution of the difficult problems in the taxonomy o f the Lupinus speeics: rf. C . L. Hitchcock a n d A . Cronquist, "Sazifragaceae t o I h c n r e a e , Vascular Plants of the Pacific Northwest," P a r t 3 , University of \\ashinuton Press, Seattle, IVasIi.. 1961, p 297 ff, 327 ff. Specimens of t h e
r.
A second crystalline alkaloid (mp 103.5-104.5') isolated by us from L. sericeus Pursh in much smaller yield than ( - )-p-isosparteine, was recognized as being one of the possible hydroxy derivatives of (-)-Pisosparteine because of the composition (C,H2&20), the presence of a strong hydroxyl band in the infrared spectrum, and a general similarity of the infrared spectra of the hydroxy base and of the major alkaloid, pisosparteine; this similarity was particularly striking in the case of the monoperchlorate salts. I n the remainder of this paper, data leading to structure I1 for thi.: alkaloid are discussed. The very unreactive character of the hydroxyl group caused us to exclude as improbable the location of the oxygen function at a position in rings A or D, since the many known hydroxysparteine derivatives having the hydroxyl qubstituents in these rings behave more or less typically as alcohols. A carbinolamine structure with the hydroxyl at any position a to a ring nitrogen (C-2, C-6, (3-10, C-11, C-13, (3-17) was also quickly ruled out, for the solutions of the hydroxy base in acids showed no tendency toward the very characteristic formation of immoniuni salts, or any of the other chemical behavior typical of a carbinolamine. Especially, no reduction occurred with lithium aluminum hydride. The chemical inertness of the hydroxyl function was exemplified by the failure of that group to form a toluenesulfonate ester under any of the usual preparative conditions, complete resistance to oxidation to form a ketonic function, resistance to dehydrating agents, acetylation only under fairly vigorou\ conditions, and reyistance to replacement of the hydroxyl group with halogen except under the special conditions of the Landauer and Rydon reagents, to be described below. Typical of the inertness of the alcohol to oxidation were the results upon treatment of the hydroxy base with (a) alkaline ferricyanide, a reagent which readily brings about oxidation of p-isosparteine to its 10,17dioxo derivative, (b) chromic anhydride-pyridine, and (e) chromic anhydride-acetic acid. ,411 of these oxidizing conditions cauied the formation from the hydroxy base of the same crystalline hydroxy lactsm (C15H24S202), mp 218.3-219" cor. It was finally possible to effect the replacement of the hydroxyl function with hydrogen by reduction with lithium aluminum hydride of the iodo derivative, which was prepared as a highly reactive compound by vigorous treatment of the hydroxy base Tvith triphenyl phosphite plant material used b j us in this in\estigation are on file in the Indiana Unnersity Herbarium under kccession No 112346-112317 ( 8 ) 3 Okuda, K Tsuda, and H Kataoka Chem Ittd (London), 1115 (1961)
3046
CARMACK, GOLDBERG, AND XARTIN
m e t h i ~ d i d e . ~The chloro derivative was prepared by an analogous procedure. I n spite of the difficulty in effecting the replacement of the hydroxyl group, t'he halogeno derivatives readily hydrolyzed back to the starting hydroxy base. The formation of the halogen substitution products was demonstrated by paper and column chromatography. The reductive elimination of halogen was carried out directly upon the reaction mixture with triphenylphosphite methiodide without separat,ion, and the reduction product, (-)-p-isosparteine, was isolated as its characteristic monoperchlorate salt ; unchanged starting mat'erial was also recovered together with t'he deoxygenated base. The chemical evidence therefore led us to conclude that the hydroxy base is either (-)-7-hydroxy-pisosparteine or ( -)-8-hydroxy-p-isosparteine. Either of these structures would be uniquely interesting from a biogenetic point of view. lo At the time the original chemical studies were carried out,'C*2 most of the published chemical information on bridgehead hydroxyl groups in small bridged systems such as apocamphanol" strongly suggested that a hydroxyl group at, the bridgehead positions, C-7 or C-9, of the sparteine nucleus would be extremely difficult or impossible to replace with a halogen substituent, and that such a bridgehead halogenated derivative, if it could be obtained by any means, would be extremely sluggish in its replacement reactions. Thus, the reactivity of the halogeno derivatives obtained by use of the Landauer and Rydon reagents was interpreted by us'c and by ot'hers? as tending to rule out the 7hydroxy-p-isosparteine structure for our alkaloid from L. sericeus. Efforts to obtain positive nmr spectral evidence supporting or eliminating the 8-hydroxy-p-isosparteine st.ructure were disappointing in that the spectra of pisosparteine and its natural hydroxy derivative, as well as selected salts, were nearly identical in the region in which a hydrogen atom [>CCH(OH)Cet ( J = 11.5 cps) with resonances centered at 2.48 and 3.05 ppm. This AB spin system is assignable to a nonequivalent methylene group.8 The identity of the spirohydantoin system was further substantiated by comparing, and finding identical, the product obtained by rearranging the alloxazinium salt 11 with the known 12, which was obtained from the ureide 13 by the procedure of Clark-Len-isga(eq 1). (7) A. R . Katritzky and A . P. dmhler in “Physical Methods in Heterocyclic Chemistry,” Vol. 11, A. R. Katritzky, Ed., Academic Press Inc., New T o r k , N. Y., 1963, p 228. (8) R . C. Cookson, T. A. Crabh, J. J. Frankel, and J. Hudec, Tetrahedron, Suppl., 7 , 355 (1966). (9) (a) J. W. Clark-Lewis, J. Chem. Soc., 422 (195i); (b) J. W. ClarkLewis and M. J. Thompson, ibid., 430 (1957); (c) F. E. King and J. W. Clark-Lewis, ibid., 3379 (1951).
